Parallel Search API

openapi: 3.2.0
info:
  title: Parallel Chat API (Beta) Chat API (Beta) Search API
  description: Parallel API
  contact:
    name: Parallel Support
    url: https://parallel.ai
    email: support@parallel.ai
  version: 0.1.2
servers:
- url: https://api.parallel.ai
  description: Parallel API
security:
- ApiKeyAuth: []
tags:
- name: Search
  description: 'Search returns ranked URLs with extended excerpts suitable for LLM consumption. Inputs are a natural-language objective and optional keyword queries. Source policies allow including or excluding specific domains and have configurable output sizes. The returned extended snippets contain dense, relevant information from relevant pages.

    - Result: ranked list with URL, title, and long text excerpts'
paths:
  /v1/search:
    post:
      tags:
      - Search
      summary: Search
      description: 'Searches the web.


        The legacy Search API reference (`/v1beta/search` endpoint) is available

        [here](https://docs.parallel.ai/api-reference/legacy/search-beta/search), and

        migration guide is [here](https://docs.parallel.ai/search/search-migration-guide).'
      operationId: v1_search_v1_search_post
      requestBody:
        content:
          application/json:
            schema:
              $ref: '#/components/schemas/V1SearchRequest'
        required: true
      responses:
        '200':
          description: Successful Response
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/V1SearchResponse'
              example:
                search_id: search_fcb2b4f3c75e418687bccaa1a8381331
                results:
                - url: https://www.example.com
                  title: Sample webpage title
                  publish_date: '2024-01-15'
                  excerpts:
                  - Sample excerpt 1
                  - Sample excerpt 2
                session_id: session_fcb2b4f3c75e418687bccaa1a8381331
        '422':
          description: Request validation error
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
              example:
                type: error
                error:
                  ref_id: search_fcb2b4f3c75e418687bccaa1a8381331
                  message: Request validation error
      x-code-samples:
      - lang: Python
        source: "from parallel import Parallel\n\nclient = Parallel()\n\nsearch = client.search(\n    objective=\"Find latest information about Parallel Web Systems. Focus on new product releases, benchmarks, or company announcements.\",\n    search_queries=[\"Parallel Web Systems products\", \"Parallel Web Systems announcements\"],\n)\nprint(search.results)"
      - lang: TypeScript
        source: "import Parallel from \"parallel-web\";\n\nconst client = new Parallel();\n\nconst search = await client.search({\n    objective: \"Find latest information about Parallel Web Systems. Focus on new product releases, benchmarks, or company announcements.\",\n    search_queries: [\"Parallel Web Systems products\", \"Parallel Web Systems announcements\"],\n});\nconsole.log(search.results);"
      - lang: cURL
        source: "curl --request POST \\\n    --url https://api.parallel.ai/v1/search \\\n    --header 'Content-Type: application/json' \\\n    --header 'x-api-key: <api-key>' \\\n    --data '{\n    \"objective\": \"Find latest information about Parallel Web Systems. Focus on new product releases, benchmarks, or company announcements.\",\n    \"search_queries\": [\"Parallel Web Systems products\", \"Parallel Web Systems announcements\"]\n}'"
components:
  schemas:
    V1SearchRequest:
      properties:
        objective:
          anyOf:
          - type: string
          - type: 'null'
          title: Objective
          description: Natural-language description of the underlying question or goal driving the search. Used together with search_queries to focus results on the most relevant content. Should be self-contained with enough context to understand the intent of the search.
        search_queries:
          items:
            type: string
          type: array
          title: Search Queries
          description: Concise keyword search queries, 3-6 words each. At least one query is required, provide 2-3 for best results. Used together with objective to focus results on the most relevant content.
        mode:
          anyOf:
          - type: string
            enum:
            - turbo
            - basic
            - advanced
          - type: 'null'
          title: Mode
          description: 'Search mode preset: supported values are `turbo`, `basic`, and `advanced`. Turbo mode is optimized for the fastest responses. Basic mode offers low latency and works best with 2-3 high-quality search_queries. Advanced mode provides higher quality with more advanced retrieval and compression. Defaults to `advanced` when omitted.'
        max_chars_total:
          anyOf:
          - type: integer
          - type: 'null'
          title: Max Chars Total
          description: Upper bound on total characters across excerpts from all results.
        session_id:
          anyOf:
          - type: string
            maxLength: 1000
          - type: 'null'
          title: Session Id
          description: Session identifier to track calls across separate search and extract calls, to be used as part of a larger task. Specifying it may give better contextual results for subsequent API calls.
        client_model:
          anyOf:
          - type: string
          - type: 'null'
          title: Client Model
          description: The model generating this request and consuming the results. Enables optimizations and tailors default settings for the model's capabilities.
          examples:
          - claude-opus-4-7
          - gpt-5.4
          - gemini-3.1-pro
        advanced_settings:
          anyOf:
          - $ref: '#/components/schemas/AdvancedSearchSettings'
          - type: 'null'
          description: Advanced configuration for source policy, fetch policy, and excerpt settings. May impact result quality and latency unless used carefully. When omitted, excerpts are enabled by default.
      additionalProperties: false
      type: object
      required:
      - search_queries
      title: V1SearchRequest
      description: Search request.
    UsageItem:
      properties:
        name:
          type: string
          title: Name
          description: Name of the SKU.
          examples:
          - sku_search_additional_results
          - sku_extract_excerpts
        count:
          type: integer
          title: Count
          description: Count of the SKU.
          examples:
          - 1
      type: object
      required:
      - name
      - count
      title: UsageItem
      description: Usage item for a single operation.
    SourcePolicy:
      properties:
        include_domains:
          items:
            type: string
          type: array
          title: Include Domains
          description: List of domains to restrict the results to. If specified, only sources from these domains will be included. Accepts plain domains (e.g., example.com, subdomain.example.gov) or bare domain extension starting with a period (e.g., .gov, .edu, .co.uk). The combined number of domains in include_domains and exclude_domains cannot exceed 200.
          examples:
          - - wikipedia.org
            - usa.gov
            - .edu
        exclude_domains:
          items:
            type: string
          type: array
          title: Exclude Domains
          description: List of domains to exclude from results. If specified, sources from these domains will be excluded. Accepts plain domains (e.g., example.com, subdomain.example.gov) or bare domain extension starting with a period (e.g., .gov, .edu, .co.uk). The combined number of domains in include_domains and exclude_domains cannot exceed 200.
          examples:
          - - reddit.com
            - x.com
            - .ai
        after_date:
          anyOf:
          - type: string
            format: date
          - type: 'null'
          title: After Date
          description: Optional start date for filtering search results. Results will be limited to content published on or after this date. Provided as an RFC 3339 date string (YYYY-MM-DD).
          examples:
          - '2024-01-01'
      type: object
      title: SourcePolicy
      description: 'Source policy for web search results.


        This policy governs which sources are allowed/disallowed in results.'
    Warning:
      properties:
        type:
          type: string
          enum:
          - spec_validation_warning
          - input_validation_warning
          - warning
          title: Type
          description: Type of warning. Note that adding new warning types is considered a backward-compatible change.
          examples:
          - spec_validation_warning
          - input_validation_warning
        message:
          type: string
          title: Message
          description: Human-readable message.
        detail:
          anyOf:
          - additionalProperties: true
            type: object
          - type: 'null'
          title: Detail
          description: Optional detail supporting the warning.
      type: object
      required:
      - type
      - message
      title: Warning
      description: Human-readable message for a task.
    V1WebSearchResult:
      properties:
        url:
          type: string
          title: Url
          description: URL associated with the search result.
        title:
          anyOf:
          - type: string
          - type: 'null'
          title: Title
          description: Title of the webpage, if available.
        publish_date:
          anyOf:
          - type: string
          - type: 'null'
          title: Publish Date
          description: Publish date of the webpage in YYYY-MM-DD format, if available.
        excerpts:
          items:
            type: string
          type: array
          title: Excerpts
          description: Relevant excerpted content from the URL, formatted as markdown.
      type: object
      required:
      - url
      - excerpts
      title: V1WebSearchResult
      description: A single search result from the web search API.
    V1SearchResponse:
      properties:
        search_id:
          type: string
          title: Search Id
          description: 'Search ID. Example: `search_cad0a6d2dec046bd95ae900527d880e7`'
        results:
          items:
            $ref: '#/components/schemas/V1WebSearchResult'
          type: array
          title: Results
          description: A list of search results, ordered by decreasing relevance.
        warnings:
          anyOf:
          - items:
              $ref: '#/components/schemas/Warning'
            type: array
          - type: 'null'
          title: Warnings
          description: Warnings for the search request, if any.
        usage:
          anyOf:
          - items:
              $ref: '#/components/schemas/UsageItem'
            type: array
          - type: 'null'
          title: Usage
          description: Usage metrics for the search request.
        session_id:
          type: string
          title: Session Id
          description: Session identifier, echoed back from the request if provided, otherwise generated by the server. Should be passed to future search and extract calls made by the agent as part of the same larger task.
          examples:
          - session_8a911eb27c7a4afaa20d0d9dc98d07c0
      type: object
      required:
      - search_id
      - results
      - session_id
      title: V1SearchResponse
      description: Search response.
    AdvancedSearchSettings:
      properties:
        source_policy:
          anyOf:
          - $ref: '#/components/schemas/SourcePolicy'
          - type: 'null'
          description: Domain and date filtering preferences for search results.
        fetch_policy:
          anyOf:
          - $ref: '#/components/schemas/FetchPolicy'
          - type: 'null'
          description: 'Fetch policy: determines when to return cached content from the index (faster) vs fetching live content (fresher). Default is to disable live fetch and return cached content from the index. Note: enabling live fetch significantly increases search latency because it requires fetching content from source websites.'
        excerpt_settings:
          anyOf:
          - $ref: '#/components/schemas/V1ExcerptSettings'
          - type: 'null'
          description: Controls excerpt sizes. Provide excerpt settings for fine-grained control, or omit to use defaults.
        location:
          anyOf:
          - type: string
          - type: 'null'
          title: Location
          description: ISO 3166-1 alpha-2 country code for geo-targeted search results.
          examples:
          - us
          - gb
          - de
          - jp
        max_results:
          anyOf:
          - type: integer
          - type: 'null'
          title: Max Results
          description: Upper bound on the number of results to return. Defaults to 10 if not provided.
      additionalProperties: false
      type: object
      title: AdvancedSearchSettings
      description: 'Advanced search configuration.


        These settings may impact result quality and latency unless used carefully.

        See https://docs.parallel.ai/search/advanced-search-settings for more info.'
    ErrorResponse:
      properties:
        type:
          type: string
          const: error
          title: Type
          description: Always 'error'.
        error:
          $ref: '#/components/schemas/Error'
          description: Error.
      type: object
      required:
      - type
      - error
      title: ErrorResponse
      description: Response object used for non-200 status codes.
    V1ExcerptSettings:
      properties:
        max_chars_per_result:
          anyOf:
          - type: integer
          - type: 'null'
          title: Max Chars Per Result
          description: Optional upper bound on the total number of characters to include per url. Excerpts may contain fewer characters than this limit to maximize relevance and token efficiency.
      additionalProperties: false
      type: object
      title: V1ExcerptSettings
      description: Optional settings for returning relevant excerpts.
    Error:
      properties:
        ref_id:
          type: string
          title: Reference ID
          description: Reference ID for the error.
        message:
          type: string
          title: Message
          description: Human-readable message.
        detail:
          anyOf:
          - additionalProperties: true
            type: object
          - type: 'null'
          title: Detail
          description: Optional detail supporting the error.
      type: object
      required:
      - ref_id
      - message
      title: Error
      description: An error message.
    FetchPolicy:
      properties:
        max_age_seconds:
          anyOf:
          - type: integer
          - type: 'null'
          title: Max Age Seconds
          description: Maximum age of cached content in seconds to trigger a live fetch. Minimum value 600 seconds (10 minutes).
          examples:
          - 86400
        timeout_seconds:
          anyOf:
          - type: number
          - type: 'null'
          title: Timeout Seconds
          description: Timeout in seconds for fetching live content if unavailable in cache.
          examples:
          - 60
        disable_cache_fallback:
          type: boolean
          title: Disable Cache Fallback
          description: If false, fallback to cached content older than max-age if live fetch fails or times out. If true, returns an error instead.
          default: false
      type: object
      title: FetchPolicy
      description: Policy for live fetching web results.
  securitySchemes:
    ApiKeyAuth:
      type: apiKey
      in: header
      name: x-api-key